Logistic regression analysis of drought persistence in East China

Lei Meng, Trenton Wayne Ford, Ying Guo

Research output: Contribution to journalArticle

Abstract

East China (EC) has experienced a number of severe droughts in recent decades. Understanding the characteristics of droughts and their persistence will provide operational guidelines for water resource management and agricultural production. This study uses a logistic regression model to measure the probability of drought occurrence in the current season given the previous season's Standardized Precipitation Index (SPI) and Southern Oscillation Index (SOI) as well as drought persistence. Results reveal large spatial and seasonal variations in the relationship between the previous season's SPI and the drought occurrence probability in a given season. The drought persistence averaged over the entire study area for all the four seasons is approximately 34% with large variations from season to season and from region to region. The East and Northeast regions have the largest summer drought persistence (∼40%) and lowest fall drought persistence (∼28%). The spatial pattern in winter and spring drought persistence is dissimilar with stronger winter and weaker spring drought persistence in the Southwest and Northeast relative to other regions. Logistic regression analysis indicates a stronger negative relationship in summer-to-fall (or between fall drought occurrence and summer SPI) than other inter-season relationships. Most of the stations with significant relationships in summer-to-fall lie in the East and the southern part of the North region. Significant negative relationships in winter–spring and fall–winter are primarily located on the western portion of the study area. A significant negative spring–summer relationship exists in the North region. Further, spring drought occurrence has a statistically significant positive relationship with winter SOI in the Southeast. Fall drought occurrence and summer SOI are negatively correlated in the northwest portion of EC (i.e. the border of North and Southwest regions). Overall, this study demonstrates that the impact of previous season SPI and SOI on current season drought varies substantially from region to region and from season to season. This study also shows stronger drought persistence in summer than in other seasons. In other words, the probability of fall drought occurrence is closely related to summer moisture conditions in the EC.

Original languageEnglish (US)
Pages (from-to)1444-1455
Number of pages12
JournalInternational Journal of Climatology
Volume37
Issue number3
DOIs
StatePublished - Mar 15 2017
Externally publishedYes

Keywords

  • drought persistence
  • East China
  • logistic regression model
  • Standardized Precipitation Index

ASJC Scopus subject areas

  • Atmospheric Science

Fingerprint Dive into the research topics of 'Logistic regression analysis of drought persistence in East China'. Together they form a unique fingerprint.

  • Cite this